Smart band heart rate sensor

In this example, we will simulate a smart band optical heart rate sensor with dynamic movement using Ansys Speos and Ansys Motion. First, the displacement of the smart band and the deformation of tissue on the human wrist will be simulated by Ansys Motion. Then, we will introduce how to import the displacement and deformation data to Speos. Finally, inside Ansys Speos, users could simulate the influence of smart band displacement and human tissue deformation on the optical signal that will be collected by the smart band heart rate sensor.

download example

Overview

Understand the simulation workflow and key results

Step 1: Smart band dynamic movement simulation in Mechanical Motion (not covered in this article)

First, we will only analyze the influence of smart band movement on the optical signal. The detailed simulation process inside Motion will not be covered in this example. Ansys Motion is an advanced engineering solution based on flexible multibody dynamics. It enables fast and accurate analysis of rigid and flexible bodies within a single solver. In Motion, we simulated a scenario where the human arm is waving orientally. During the arm movement, the smart band slides accordingly. The relative displacement between the smart band and the human wrist will be exported to Speos.

Step 2: Build up tissue model and simulation in Speos

In the second step, we build up the wrist structure with optical parameters in Speos. The tissue is modeled according to the model described in the following knowledge base article ” How to model the human skin and optical heart rate sensors in OpticStudio ”. As introduced in this article, the sub-layers of the dermis are modeled separately since they have different blood content values. The epidermis is modeled as one thick layer as there is no blood inside. The refraction, absorption and scattering properties of tissue are considered in this model.

Step 3: Speos batch Simulation with Workbench

The relative displacement of the smart band to the human wrist is exported from motion. The displacement of the smart band as a function of time in X, Y, and Z directions is saved in three .CSV tables separately. The design of the experiment (DOE) is built with Workbench. The variable is the displacement of the smart band and the output would be the irradiance that is received by the smart band sensor. The influence of smart band movement on the received optical signal is then analyzed.

Step 4: Tissue deformation simulation in Mechanical Motion (not covered in this article)

In this step, we will simulate the deformation of tissue when pushing the smart band toward human wrist tissue. The detailed simulation process of tissue deformation is not covered in this article. Motion could output the deformed model. Here we chose to export .STL files of each deformed tissue layer.

Step 5: Import the deformed tissue structure into Speos, then launch the simulation

The deformed tissue model is imported to Speos. The optical properties are then applied to corresponding tissue layers. By this simulation, the user could analyze the influence of tissue deformation on the retro-scattered signals. Users could analyze and compare the received signal between a tightly wared band and a loosely wared one.

Run and Results

Instructions for running the model and discussion of key results

Step 1: Smart band dynamic movement simulation in Mechanical Motion (not covered in this article)

The simulation process in Motion is not covered in this article. To simplify the problem, in the simulation of the smart band displacement, the human wrist is modeled as a rigid body. The deformation of the tissue is ignored in this step. From Ansys Motion simulation, the relative displacement of the smart band could be exported. In “**\Ansys SmartBand 23R1\Motion_displacement”, the user could find “DispX.csv” “DispY.csv” and “DispZ.csv”. These three tables contain respectively the smart band displacement data in X, Y, and Z directions as a function of time. This data will be used as the design points in the workbench design of the experiment.

Step 2: Build up tissue model and simulation in Speos

1. Open Ansys SmartBand 23R1.scdocx
2. Run direct.1 simulation.
3. Open Direct.1.Irradiance.xmp
4. Open the measurement tool then maximize the measurement area.
5. Click the file then select export template to export the measurement template.
   Direct.1.Irradiance.xml

In this step, we need to prepare the smart band and human wrist model and set up the optical simulation inside Ansys Speos. The biological structure of the human wrist and the optical properties of each biological layer have been built according to the one introduced in the Zemax knowledgebase article. The LED light source could be modeled as a Lambertian surface source. Its output power has been normalized as 1 W to simplify the analysis. The source spectrum is centered at 525 nm which is the most used wavelength for heart rate monitors. The optical sensor is simplified as an irradiance sensor.

After opening the project, the user could find the built-up human wrist and smart band 3D model. The layered structures in the human wrist and their optical properties have been set. After going through the introduced 6 steps, the user could get the following result.

Step 3: Speos batch Simulation with Workbench

To make the batch Simulation with the Workbench, the user needs to first create the input variable inside Speos. As mentioned above, the output of the Motion simulation is the relative movement of the smart band in the X, Y, and Z directions. Consequently, the user needs to define the corresponding parameters inside Speos.

1. In the group section, create three script parameters X, Y, and Z. These parameters are defined as the relative movement parameters in the corresponding directions (as shown below).
2. Create a Script group. Right-click on the created group (Group1 for example), and select edit script. Then click the record button inside the script window.
3. Select the Move function under the Design tab. Select the “Smart Watch” component inside the structure tree. The process to record the necessary code is shown in the following gif.
4. After finishing the previous steps, the user has prepared the input and output parameters for the design of the experiment inside Ansys Workbench. The precise steps to build a design of an experiment inside Workbench are not covered in this article. You may find the built-in project in Ansys SmartBand 23R1.wbpj.
5. Open Ansys SmartBand 23R1.wbpj then double click on Parameter sets.
6. Input the relative displacement data in X Y and Z directions respectively.
7. Click the Update All Design point button. The simulation results at every design point will be calculated automatically. Results can be found under the Ansys SmartBand 23R1_files folder

In the GIF above, the Move control handle is positioned at the axis of the irradiance sensor. This step is not necessary, because the exported data from Motion is the relative displacement instead of the Global coordinate, therefore the control handle can be attached at any preferred location. After recording the Move commands, X, Y, and Z parameters are added to the target coordinate command line so that the position of the smart band can be controlled as shown in the gif below. The parameters and the script have been included in the project.

Step 4: Tissue deformation simulation in Ansys Motion

The process to simulate the tissue deformation is out of the scope of this article. But we need to mention that the 2D simulation is selected instead of full 3D deformation simulation in Motion. As we introduced in the previous step, all layers that contain blood need to be simulated to achieve a convincing result. But the complex layered structure will dramatically increase the calculation loads. To accelerate the calculation speed, only the deformation in XY plane is considered. The deformation in Z direction is supposed to be constant. In the 2D simulation the smart band is pushed 0.92 mm into the wrist. Users could find the deformed layer model under \deformation\2D_Result.

Step 5: Import the deformed tissue structure into Speos, then launch the simulation

The 2D deformed layers are imported to Speos. The 2D faces are pulled in the Z direction to reconstruct the 3D structure. Users could find the reconstructed 3D model in “\deformation\deformation.scdocx”.

1. Open deformation.scdocx
2. Run Direct.1 simulation

Here is the comparison between the signal received by the sensor before and after pushing the smart band towards the wrist (left image: no deformation, right image: with deformation). The total flux is 0.36 lm (left image) compared to 1.83 lm (right image). The simulation result conforms to our intuitive guess, when tightening up the smart band, its PPG sensor could have a signal reception.

Important Model Settings

Description of important objects and settings used in this model

Motion data export and setup in Workbench

In the demo project, the design points are the relative movements at each second. It should be noticed that, in this demo the definition of displacement directions inside Motion and Speos are inverse. Therefore, the data in Workbench represents the inverse values of the Motion outputs. The 'Y' direction denotes vertical movement. In the Motion simulation of smart band movement, the human tissue and smart band are simplified as rigid bodies, vertical displacement should consistently register as positive (otherwise it means the smart band model penetrates the tissue model).

However, due to simulation precision, minor negative values may occasionally arise, typically on the order of magnitude less than 1E-4 in this demo. These small negative values are disregarded and replaced with zero during the setup of the design of the experiment.

Additional resources

Additional documentation, examples, and training material

See also

Relevant Ansys Learning Hub courses on Speos:

• Ansys Speos Getting Started
• Ansys Speos Workflow and Geometry Management
• Ansys Workbench Table of Design Points

Related Publications

1. Maeda, et al. Monte Carlo simulation of spectral reflectance using a multilayered skin tissue model. Optical Review (2010)
2. Sinichkin, et al. In vivo fluorescence spectroscopy of the human skin: experiments and models. Journal of Biomedical Optics (1998)
3. Gae Hwang Lee. Stretchable PPG sensor with light polarization for physical activity–permissible monitoring. Science Advances (2022)